An Activated Sludge Plant Can Hold Its F/M Ratio Steady and Still Lose Nitrification Underneath It

Most operators running an activated sludge process, the biological core of most ETPs and STPs built around this process, track the food-to-microorganism ratio, the daily BOD load divided by the mixed liquor suspended solids inventory in the aeration tank, as the day-to-day health check on the biology. It is a reasonable number to watch and it is easy to compute from readings a plant already takes. It is not, however, the number that determines whether the slow-growing nitrifying bacteria the plant depends on for ammonia removal can actually survive in the system, because that depends on sludge retention time, commonly called sludge age, and SRT can drift away from a stable F/M ratio without either number ever looking obviously wrong on its own.

Two different ratios, answering two different questions

F/M ratio is a load-normalized snapshot: how much food is arriving today relative to how much biomass is currently in the tank. It answers whether the system is under- or over-loaded relative to its biomass inventory right now. SRT answers a completely different question: on average, how many days does a unit of biomass stay in the system before it is wasted out. That second number is what actually determines whether the bacteria carrying out nitrification, whose growth rate is far slower than the heterotrophic bacteria that dominate ordinary BOD removal, can reproduce fast enough to replace what is wasted out before they are washed out of the system entirely. A system commonly needs an SRT in the range of eight days or more to sustain reliable nitrification, well above what BOD removal alone would require.

Why the two numbers can move independently

F/M ratio is computed from flow and load, both of which change hour to hour and day to day for reasons that have nothing to do with how the plant is being operated: weather, industrial discharge patterns, seasonal variation in influent strength. SRT is set by a different decision entirely, how much sludge the operator wastes out, usually on a schedule or a target MLSS setpoint rather than recalculated against F/M in real time. A plant can hold F/M inside its normal target band for weeks by virtue of flow and load staying roughly where they usually are, while the wasting rate behind SRT has quietly drifted, whether from a wasting pump running slightly under its set rate, a schedule that was never revisited after a process change, or simply RAS and WAS being adjusted together without anyone recalculating what that combination now implies for sludge age. Neither number's own trend chart would show the other number's problem, because they are driven by different inputs entirely.

This is the same shape of mistake we described in the piece on motor insulation resistance trending: a number that is checked correctly and reads fine on its own terms can still be sitting on top of a second, independently moving number that the first number's own chart was never built to reveal. There, the checked number was an annual pass-fail spot reading sitting on top of an unplotted multi-year trend. Here, it is F/M ratio sitting on top of an SRT that depends on an entirely separate control decision. In both cases the operator is watching a real, correctly computed number that simply is not the number the failure mode actually depends on.

Picture a plant that nitrifies reliably at commissioning, with F/M ratio and SRT both landing where the design intended. Over the following year, incremental changes accumulate the way they do at most plants: the wasting rate gets nudged down slightly to manage sludge handling costs, a new side stream gets routed into the aeration basin changing the load balance, and nobody revisits the SRT calculation because F/M, driven mostly by flow and BOD load rather than by the wasting decision, keeps reading inside its usual range the whole time. SRT, meanwhile, has drifted down from ten days to six, below what the plant's nitrifier population needs to sustain itself. Ammonia removal does not fail gradually and visibly the way a slowly rising cake moisture does; it tends to hold until the nitrifier population falls below a threshold and then degrades fast, often coinciding with a cold-weather growth-rate dip that exposes the margin had already disappeared.

By the time ammonia in the final effluent crosses a consent limit, the operator reviewing the trend charts usually finds an F/M ratio that looks exactly like it always has, because the flow and load driving it never did anything unusual. The wasting log, if anyone pulls it and reconstructs SRT month by month, tells the real story: a slow, steady decline that nothing in the plant's routine reporting was built to surface, because F/M was the number everyone was already watching and F/M never moved. The gap between what was being checked and what was actually failing is the entire problem, not a measurement error or an unusually fast failure.

What tracking SRT separately actually requires

The fix is not replacing F/M monitoring, which is still a legitimate load-balance check. It is calculating and tracking SRT as its own number, on its own schedule, rather than assuming a stable F/M implies a stable sludge age. SRT is computed from the mixed liquor solids inventory divided by the solids wasted per day, a calculation most plants already have the underlying data for in their daily wasting and MLSS records, it simply is not always performed and trended as a distinct metric the way F/M usually is. Plants running nutrient removal processes in particular are the ones with the least margin for this drift to go unnoticed, since their design SRT already sits close to what nitrifiers minimally require, leaving little room before a slow wasting-rate drift turns into a visible ammonia exceedance.

None of this requires new instrumentation. A plant that already logs MLSS and daily waste activated sludge volume has every input SRT needs; the only missing step is calculating and trending it separately rather than treating a stable F/M ratio as proof that sludge age is stable too. A monthly SRT calculation, checked against the plant's own nitrification requirement rather than inferred from F/M, catches a slow wasting-rate drift months before it shows up as an ammonia violation.

If your plant tracks F/M ratio but has not separately calculated sludge retention time in a while, that gap is worth closing with Spans' process engineering team before a quiet wasting-rate drift turns into a nitrification failure nobody saw building.